Hacker News Reader

Rather than viewing it as “trustless” you can view it as “trust minimized”. Then it might make more sense. Trust with your counterparty falls on a spectrum.

You trust your friend, but maybe not an anonymous online individual. So you might use the blockchain to accept payments from the anonymous individual whereas with your friend, you might accept a check or Venmo.

The only situation where trust is minimized or absent is jurisdictions where there is no law or less law, in which case you’re going to use force instead of code to enforce law.

Trust is an inherent property of a functioning society, and hence why the bonafide use cases for blockchain and crypto are so few. If you live somewhere where the police, a judge, or government can make you whole in a dispute, not a lot of need for web3. If you live somewhere where that isn’t the case (failing nation states, authoritarian regimes), that’s a better product market fit for web3.

And in about four hours I'm going to hear an equally-confident HN lecture about how "it's not worth it to sue" for less than $1000.

Crypto-skeptics talk like the legal system is currently equipped to robustly make you whole in those transactions. It is manifestly not. Those transactions do, in fact, depend on other means of establishing trust.

To be sure, a legal system is an important mechanism for facilitating trust. But also, be careful not to overstate its relevance either.

> If you live somewhere where the police, a judge, or government can make you whole in a dispute

You're missing the point. If you create a system where you won't need to make use of the legal system because transactions execute correctly by default, your transaction costs are suddenly way lower.

Obviously your counterparty can still sue you after the fact, but thats a much better position to be in than needing the legal system to push the transaction through in the first place.

> transactions execute correctly by default

That's not what you get, because in practice you want some physical good for your BTC and a blockchain cannot enforce that you do get it, it can only enforce that the BTC are transferred from one address to another.

Not necessarily..tons of financial transactions are purely digital (currency exchange, stocks, derivatives etc).

You can also tokenize a real world asset (for example a token representing an amount of gold). Then you have to trust the issuer of the token, but it's very possible you trust that issuer but don't trust your counterparty.

The examples you mentioned are not purely digital since they rely heavily on the legal system. You don't consider tokens abstracted from the whole physical world.

I meant that nothing is fully embedded within a blockchain (maybe physical was the wrong word).

For sure, something like a stock still has a real-life counterparty that you can't get rid of. Trading it on a blockchain eliminates only the counterparty risk from the person you trade with, not with the party issuing the token / asset. This is true for basically everything except the native token.

> Trading it on a blockchain eliminates only the counterparty risk from the person you trade with

I think we agree here. The question then is: how useful is this really?

For example, assume I have some USD that I want to trade for VT, and I want to decide between using a blockchain or not.

The counterparty risk solved by using a blockchain is:

1. No blockchain: the broker (who's between me and the person I want to trade with) can technically keep the USD and the VT.

2. Blockchain: this cannot happen because a smart contract prevents it.

But:

a. How often are you in a situation where you trust the entities backing the two assets (USD and VT here) but can't find a broker whose trustworthiness is implied by the trustworthiness of the two entities backing the two other assets?

b. Less trust is always better, but blockchain-based solutions bring their own problems (e.g. private key custody). Are there even cases in which this trade-off is worth it?

What if I:

- Am not in a jurisdiction where there is a trustworthy broker for VT or USD?

- Want to loan someone my VT (for example for them to short)?

- Want to buy some derivative of VT (easy to do in the US for VT American-style options, but this isn't true for all derivatives on all assets in all markets)

- Need to buy a large amount of VT and using the open market isn't cost effective?

VT maybe isn't the best example in favor of crypto as it's probably already one of the easiest assets in the world to trade safely online. There's lots of other stuff that's way trickier in the current system.

1. that's unlikely you're allowed to trade VT at all in this case.

2. the real hurdle is regulations rather than anything technical.

3. again, you don't need a technical solution, you need a legal one.

4. and this will be more effective thanks to a blockchain? How so? So far, blockchain-based solutions seem more costly rather than less.

The current system has a lot of limitations, but many (most?) of them are of regulatory rather than technical nature, and blockchains do not help in that regard.

> The only situation where trust is minimized or absent is jurisdictions where there is no law or less law, in which case you’re going to use force instead of code to enforce law.

Or where you don't want to share your identity... like maybe the internet?

In which case, you violate nation state laws against KYC and AML, and the repercussions that go along with that.

Having the ability to implement a solution technically does not make it legal or without consequence (like Virgil Griffith, formerly of the Ethereum Foundation, who demonstrated to North Korea how to evade sanctions trading Ethereum and is going to jail for five years for doing so).

> In which case, you violate nation state laws against KYC and AML, and the repercussions that go along with that.

No, actually. Many people send crypto transactions all the time, and do so anonymously, and the vast majority of them don't get in trouble for it.

Descriptively, it is untrue that there is some sort of wide scale crack down on anonymous crypto transactions in the west.

And no, before you say it, it doesn't really matter if, hypothetically, a future government could become extremely authoritarian, and arrest all the crypto users, because for whatever reason, crypto has been around for a decade, and governments aren't doing that.

The reality, and facts on the ground, are that people making anonymous crypto transactions all the time, and don't get in trouble with that, no matter how much people will then respond to this post by saying "Well, what if the government started doing it tomorrow!". Its not happening. Therefore, it is a useful usecase.

> Descriptively, it is untrue that there is some sort of wide scale crack down on anonymous crypto transactions in the west.

Europe: https://www.ecb.europa.eu/press/key/date/2022/html/ecb.sp220...

US Executive Order: https://www.whitehouse.gov/briefing-room/presidential-action...

UK: https://www.cnbc.com/2022/04/04/uk-to-mint-its-own-nft-and-p...

China: https://techcrunch.com/2021/09/24/china-says-all-cryptocurre...

South Korea: https://www.bbc.com/news/business-42784384

Japan: https://www.fxempire.com/news/article/financial-services-age...

Caveat being India and Japan, these are the top 10 economies in the world.

C'mon. Stop. You know that the government is not going around raiding crypto meetups and arresting people.

People send crypto transactions all the time and don't get arrested.

I know this. You know this. We don't have to pretend.

People use crypto all the time and don't get in trouble for it. We both know this.

I used crypto just this morning on chain. Nothing happened to me, even though I didn't send a KYC letter to the government about it.

The government isn't going to show up at my doorstep tomorrow because I used a smart contract.

Your feelings, are what is called "cope". People have been predicting for years and years that all the crypto people were going to be sent to jail, yet it keeps not happening.

Instead, the only thing that has happened, is that more and more of the entire tech industry has moved towards using more crypto.

I went to GDC a couple weeks ago, and I learned that a freaking 3rd of the video games industry is desperately trying to figure out how to add NFTs to their games.

I wonder what industry will be next? And will it ever been enough, to dismiss the constantly wrong people, who continue to be wrong about all the crypto people going to jail? Because I remember the same arguments a decade ago. Still hasn't happened yet.

I still go to crypto meetups every week, and I am not seeing the police raids, for all the illegal things that you think we are doing.

Or, in other words. Cope.

The legal system is slow

Slow, unreliable, and expensive. I think mostly people in the U.S. significantly overestimate how much access they actually have to the legal system. This misconception vanishes very quickly the first time one has occasion to interact with the legal system.

It's easier to have trust when there's less trust required, that statement should make sense without seeing it as an attack on inherent properties of a functioning society. And the reason that the financial system didn't break down entirely in 2008 is that taxpayers made the banks whole again. Nothing has changed fundamentally about the way banking works since, they've just added some regulations that reduce the extent of the same activities that led to the crisis.

> If you live somewhere where the police (...) can make you whole in a dispute, not a lot of need for web3. (...) somewhere where that isn’t the case (...), that’s a better product market fit for web3.

Being part of a "functioning society" is not a static property and even less so in this globalized world.

Try coordinating work with a team distributed around the world and you will see that your functioning institutions do not help your colleague in Venezuela, Argentina, Syria or Ukraine. And if you don't have/know people from these places, ask yourself why. Try helping a Brazilian immigrant living in the US that wants to get their savings to buy property back home, see how painful/inefficient it is to send anything over $2-3 thousand. Suddenly your appreciation of "web3" increases.

And even just in your own developed country, there is no guarantee that things will stay functional forever. Just look at the US and the events of January, Hong-Kong, or any country that was a former state of the USSR and needs to worry about Vlad.

A decentralized web that does not depend on the functional institutions is insurance: you pray that you will never use it, but you should be glad to know it is there if you need it.

While I suppose I technically agree with you that you can't literally remove all trust by some definitions, that doesn't mean it's not vitally important to talking about and carefully considering who you trust and in many cases use technology to remove the reliance on trusting particular parties.

For example, if I say "I want to use TLS to order something from an online retailer because I don't want to trust my ISP with my credit card and purchase information" it would not be appropriate to respond with "well that's just ridiculous, because there's no way to remove all trust, after all you still have to trust the online retailer to ship you the item, you have to trust your credit card company to not overcharge you, etc."

You're correct but that's not the way that these systems are promoted. I wouldn't have any problem with that particular aspect if the trust issue was framed that way, and there was some kind of real threat modeling going on. The idea of cryptocurrency is still terrible however.

Note your quote from the author was talking about collaboration without requiring trust in collaborators not a system without requiring trust in the system. 51% attacks are the closest to what the author was actually talking about but that has more to do with trusting and accepting how the system (and as you say, somewhat reality) works than any particular trust in a collaborator.

That said it still doesn't make the list of cases this capability is needed any longer.

Don't fall into the trap of word-thinking. When people talk about "trustless" systems, it just means that there is no hidden state. What one sees as the consensus is valid for everyone.

"Word-thinking". I like that phrase: when you over-index on the specific word as opposed to the meaning intended. For instance, thinking "act of god" is meaningless in a legal document etc.

Takk takk

“Useless” is a strong statement which probably goes too far.

One of the strongest usecases for blockchain is a distributed ledger among mostly trusted entities who frequently exchange resources.

One or more of the big banks is developing blockchain to facilitate payments between banks who end up settling enormous amounts of money between them on a daily basis. It is just nicer and much lower overhead to have a distributed and cryptographically proven way to interact with this ledger than having one bank administer it. The mostly trusted nature of the relationships between them also means if there is an error, transactions can easily be reversed or fixed and there’s no concept of anything being stolen because the ledger doesn’t actually store any value.

ScPrime and Storj use the blockchain to reduce micro-payment fees to data centers for data storage and bandwidth. Sure this could be done via traditional banking arrangements but the transaction fees would take a major bite out of the data center's profits.

No safeguard is foolproof, including protections against 51% attacks. That doesn't make the precautions pointless though. It's like saying locks are useless because they can be picked.

Sure but every blockchain I've seen make an effort to be decentralized has its "precautions" mostly in the form of an NGO that tries to manage the competing interests of every participant through various methods, to varying levels of success. A lot of them look suspiciously like unregulated credit unions or trust funds. There's that word "trust" again...

Including Bitcoin and Ethereum? Maybe I'm not sure exactly what you mean.

From what I’ve seen from LockPickingLawyer, I now think locks are only useful to help honest people know what not to take: https://youtu.be/ULUz4u5FLYg

An interesting point from the LockPickingLawyer that I'd like to point out is that while many, many locks are easy to break if you're interested, it's still effort. Your average locker or bike thief isn't picking locks, they're just taking whatever's easiest.

I think that perspective is overblown. Yeah, anyone with the inclination to practice a bit can bypass a lock. You're not going to stop a professional burglary ring with any kind of lock.

But there are a lot more lazy crackheads and dumb teenagers than professional burglars.

Is anybody else using blockchains as compute nodes? Like lambda that your users pay for to update the state of your application instead of you, while reading is completely free? I haven't seen an equivalent system thats so economical for developers.

I don't care that my application doesn't need to be on all 7000 nodes, not even a factor.

And as far as funnels go, in Web 2.0 logic the point of the "funnel" was to convert a user into a paying user, in Web 3 they are already paying users at the beginning of the funnel, and pay to update the state of your application before you ask them to do anything, this is heavily optimized for driving revenue when applying the funnel concept. So if you design your application for this, you can strip out most of the complicated frontend and backend stack and don't have to pay subscriptions to any SaaS service, and get revenue because your funnel is 1 step too.

"I don't care that my application doesn't need to be on all 7000 nodes, not even a factor."

This is an interesting topic, blockchain security is very useful for a bunch of applications (i'd prefence that to say stop trying to touch the real world with the blockchain, and you realize it's fucking amazing) but you don't need the same amount of distributed security for all applications. Banking apps need high security, gaming apps need a lot less. Avalanche is really interesting in this respect. It has the main C-Chain which has thousands of validators but it also has subnets which might have just a few validators. This gives you the ability to toggle your security needs to your application (it also let's you toggle the VM, so you do things like add stateful precompiles which let you do things that may not be practical in the EVM.

We may not be able to go completely "without trust", but we can signifiantly reduce the amount of trust required. For financial contracts that can actually make a lot of sense.

There's an army of banking regulators in every country, and yet we still see bank failures where people lose money every year (even in developed countries with functioning deposit insurance there's usually a limit on the amount that is insured), plus we get the occasional financial crisis every few decades, after the memory of the last one has faded.

Absolute statments usually don't hold up well.

Have you considered:

Settlement is much quicker. Did you know it could take upto 7 years to settle a cheque/check.

You don't have to trust the business as much as the business has to trust you before they send the goods. Chargebacks and unfavorable refunds can eat into profit. Quick settlements elimate this issue.

Why are you using the worst case scenario? Plenty of banks settle a check in a single business day.

> Settlement is much quicker. Did you know it could take upto 7 years to settle a cheque/check

That has nothing to do with blockchain and does not mean at all it is the best way to solve the problem

> Did you know it could take upto 7 years to settle a cheque/check.

The solution to cheques isn't blockchain, but simply to advance out of 19th century banking. Even a 20 year old mainframe would be an upgrade over cheques.

> Absolute statments usually don't hold up well.

There you go. Otherwise I would not see YC's most valuable startup ignoring all these critics and reconsidering and using it for Crypto Payouts if it was so-called 'unnecessary' or 'useless' in the first place. [0]

That is the problem with statements that take a very absolutist direction. It goes either way for anti-crypto supporters and also crypto supporters.

> Chargebacks and unfavorable refunds can eat into profit. Quick settlements elimate this issue.

The customer can instead contact the merchant for a refund instead of abusing the disputes system with chargeback fraud.

[0] https://stripe.com/blog/expanding-global-payouts-with-crypto